xref: /openbmc/linux/drivers/iommu/intel/dmar.c (revision 78091edc)
1 // SPDX-License-Identifier: GPL-2.0-only
2 /*
3  * Copyright (c) 2006, Intel Corporation.
4  *
5  * Copyright (C) 2006-2008 Intel Corporation
6  * Author: Ashok Raj <ashok.raj@intel.com>
7  * Author: Shaohua Li <shaohua.li@intel.com>
8  * Author: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
9  *
10  * This file implements early detection/parsing of Remapping Devices
11  * reported to OS through BIOS via DMA remapping reporting (DMAR) ACPI
12  * tables.
13  *
14  * These routines are used by both DMA-remapping and Interrupt-remapping
15  */
16 
17 #define pr_fmt(fmt)     "DMAR: " fmt
18 
19 #include <linux/pci.h>
20 #include <linux/dmar.h>
21 #include <linux/iova.h>
22 #include <linux/timer.h>
23 #include <linux/irq.h>
24 #include <linux/interrupt.h>
25 #include <linux/tboot.h>
26 #include <linux/dmi.h>
27 #include <linux/slab.h>
28 #include <linux/iommu.h>
29 #include <linux/numa.h>
30 #include <linux/limits.h>
31 #include <asm/irq_remapping.h>
32 
33 #include "iommu.h"
34 #include "../irq_remapping.h"
35 #include "perf.h"
36 #include "trace.h"
37 
38 typedef int (*dmar_res_handler_t)(struct acpi_dmar_header *, void *);
39 struct dmar_res_callback {
40 	dmar_res_handler_t	cb[ACPI_DMAR_TYPE_RESERVED];
41 	void			*arg[ACPI_DMAR_TYPE_RESERVED];
42 	bool			ignore_unhandled;
43 	bool			print_entry;
44 };
45 
46 /*
47  * Assumptions:
48  * 1) The hotplug framework guarentees that DMAR unit will be hot-added
49  *    before IO devices managed by that unit.
50  * 2) The hotplug framework guarantees that DMAR unit will be hot-removed
51  *    after IO devices managed by that unit.
52  * 3) Hotplug events are rare.
53  *
54  * Locking rules for DMA and interrupt remapping related global data structures:
55  * 1) Use dmar_global_lock in process context
56  * 2) Use RCU in interrupt context
57  */
58 DECLARE_RWSEM(dmar_global_lock);
59 LIST_HEAD(dmar_drhd_units);
60 
61 struct acpi_table_header * __initdata dmar_tbl;
62 static int dmar_dev_scope_status = 1;
63 static DEFINE_IDA(dmar_seq_ids);
64 
65 static int alloc_iommu(struct dmar_drhd_unit *drhd);
66 static void free_iommu(struct intel_iommu *iommu);
67 
68 static void dmar_register_drhd_unit(struct dmar_drhd_unit *drhd)
69 {
70 	/*
71 	 * add INCLUDE_ALL at the tail, so scan the list will find it at
72 	 * the very end.
73 	 */
74 	if (drhd->include_all)
75 		list_add_tail_rcu(&drhd->list, &dmar_drhd_units);
76 	else
77 		list_add_rcu(&drhd->list, &dmar_drhd_units);
78 }
79 
80 void *dmar_alloc_dev_scope(void *start, void *end, int *cnt)
81 {
82 	struct acpi_dmar_device_scope *scope;
83 
84 	*cnt = 0;
85 	while (start < end) {
86 		scope = start;
87 		if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_NAMESPACE ||
88 		    scope->entry_type == ACPI_DMAR_SCOPE_TYPE_ENDPOINT ||
89 		    scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE)
90 			(*cnt)++;
91 		else if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_IOAPIC &&
92 			scope->entry_type != ACPI_DMAR_SCOPE_TYPE_HPET) {
93 			pr_warn("Unsupported device scope\n");
94 		}
95 		start += scope->length;
96 	}
97 	if (*cnt == 0)
98 		return NULL;
99 
100 	return kcalloc(*cnt, sizeof(struct dmar_dev_scope), GFP_KERNEL);
101 }
102 
103 void dmar_free_dev_scope(struct dmar_dev_scope **devices, int *cnt)
104 {
105 	int i;
106 	struct device *tmp_dev;
107 
108 	if (*devices && *cnt) {
109 		for_each_active_dev_scope(*devices, *cnt, i, tmp_dev)
110 			put_device(tmp_dev);
111 		kfree(*devices);
112 	}
113 
114 	*devices = NULL;
115 	*cnt = 0;
116 }
117 
118 /* Optimize out kzalloc()/kfree() for normal cases */
119 static char dmar_pci_notify_info_buf[64];
120 
121 static struct dmar_pci_notify_info *
122 dmar_alloc_pci_notify_info(struct pci_dev *dev, unsigned long event)
123 {
124 	int level = 0;
125 	size_t size;
126 	struct pci_dev *tmp;
127 	struct dmar_pci_notify_info *info;
128 
129 	BUG_ON(dev->is_virtfn);
130 
131 	/*
132 	 * Ignore devices that have a domain number higher than what can
133 	 * be looked up in DMAR, e.g. VMD subdevices with domain 0x10000
134 	 */
135 	if (pci_domain_nr(dev->bus) > U16_MAX)
136 		return NULL;
137 
138 	/* Only generate path[] for device addition event */
139 	if (event == BUS_NOTIFY_ADD_DEVICE)
140 		for (tmp = dev; tmp; tmp = tmp->bus->self)
141 			level++;
142 
143 	size = struct_size(info, path, level);
144 	if (size <= sizeof(dmar_pci_notify_info_buf)) {
145 		info = (struct dmar_pci_notify_info *)dmar_pci_notify_info_buf;
146 	} else {
147 		info = kzalloc(size, GFP_KERNEL);
148 		if (!info) {
149 			if (dmar_dev_scope_status == 0)
150 				dmar_dev_scope_status = -ENOMEM;
151 			return NULL;
152 		}
153 	}
154 
155 	info->event = event;
156 	info->dev = dev;
157 	info->seg = pci_domain_nr(dev->bus);
158 	info->level = level;
159 	if (event == BUS_NOTIFY_ADD_DEVICE) {
160 		for (tmp = dev; tmp; tmp = tmp->bus->self) {
161 			level--;
162 			info->path[level].bus = tmp->bus->number;
163 			info->path[level].device = PCI_SLOT(tmp->devfn);
164 			info->path[level].function = PCI_FUNC(tmp->devfn);
165 			if (pci_is_root_bus(tmp->bus))
166 				info->bus = tmp->bus->number;
167 		}
168 	}
169 
170 	return info;
171 }
172 
173 static inline void dmar_free_pci_notify_info(struct dmar_pci_notify_info *info)
174 {
175 	if ((void *)info != dmar_pci_notify_info_buf)
176 		kfree(info);
177 }
178 
179 static bool dmar_match_pci_path(struct dmar_pci_notify_info *info, int bus,
180 				struct acpi_dmar_pci_path *path, int count)
181 {
182 	int i;
183 
184 	if (info->bus != bus)
185 		goto fallback;
186 	if (info->level != count)
187 		goto fallback;
188 
189 	for (i = 0; i < count; i++) {
190 		if (path[i].device != info->path[i].device ||
191 		    path[i].function != info->path[i].function)
192 			goto fallback;
193 	}
194 
195 	return true;
196 
197 fallback:
198 
199 	if (count != 1)
200 		return false;
201 
202 	i = info->level - 1;
203 	if (bus              == info->path[i].bus &&
204 	    path[0].device   == info->path[i].device &&
205 	    path[0].function == info->path[i].function) {
206 		pr_info(FW_BUG "RMRR entry for device %02x:%02x.%x is broken - applying workaround\n",
207 			bus, path[0].device, path[0].function);
208 		return true;
209 	}
210 
211 	return false;
212 }
213 
214 /* Return: > 0 if match found, 0 if no match found, < 0 if error happens */
215 int dmar_insert_dev_scope(struct dmar_pci_notify_info *info,
216 			  void *start, void*end, u16 segment,
217 			  struct dmar_dev_scope *devices,
218 			  int devices_cnt)
219 {
220 	int i, level;
221 	struct device *tmp, *dev = &info->dev->dev;
222 	struct acpi_dmar_device_scope *scope;
223 	struct acpi_dmar_pci_path *path;
224 
225 	if (segment != info->seg)
226 		return 0;
227 
228 	for (; start < end; start += scope->length) {
229 		scope = start;
230 		if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_ENDPOINT &&
231 		    scope->entry_type != ACPI_DMAR_SCOPE_TYPE_BRIDGE)
232 			continue;
233 
234 		path = (struct acpi_dmar_pci_path *)(scope + 1);
235 		level = (scope->length - sizeof(*scope)) / sizeof(*path);
236 		if (!dmar_match_pci_path(info, scope->bus, path, level))
237 			continue;
238 
239 		/*
240 		 * We expect devices with endpoint scope to have normal PCI
241 		 * headers, and devices with bridge scope to have bridge PCI
242 		 * headers.  However PCI NTB devices may be listed in the
243 		 * DMAR table with bridge scope, even though they have a
244 		 * normal PCI header.  NTB devices are identified by class
245 		 * "BRIDGE_OTHER" (0680h) - we don't declare a socpe mismatch
246 		 * for this special case.
247 		 */
248 		if ((scope->entry_type == ACPI_DMAR_SCOPE_TYPE_ENDPOINT &&
249 		     info->dev->hdr_type != PCI_HEADER_TYPE_NORMAL) ||
250 		    (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE &&
251 		     (info->dev->hdr_type == PCI_HEADER_TYPE_NORMAL &&
252 		      info->dev->class >> 16 != PCI_BASE_CLASS_BRIDGE))) {
253 			pr_warn("Device scope type does not match for %s\n",
254 				pci_name(info->dev));
255 			return -EINVAL;
256 		}
257 
258 		for_each_dev_scope(devices, devices_cnt, i, tmp)
259 			if (tmp == NULL) {
260 				devices[i].bus = info->dev->bus->number;
261 				devices[i].devfn = info->dev->devfn;
262 				rcu_assign_pointer(devices[i].dev,
263 						   get_device(dev));
264 				return 1;
265 			}
266 		BUG_ON(i >= devices_cnt);
267 	}
268 
269 	return 0;
270 }
271 
272 int dmar_remove_dev_scope(struct dmar_pci_notify_info *info, u16 segment,
273 			  struct dmar_dev_scope *devices, int count)
274 {
275 	int index;
276 	struct device *tmp;
277 
278 	if (info->seg != segment)
279 		return 0;
280 
281 	for_each_active_dev_scope(devices, count, index, tmp)
282 		if (tmp == &info->dev->dev) {
283 			RCU_INIT_POINTER(devices[index].dev, NULL);
284 			synchronize_rcu();
285 			put_device(tmp);
286 			return 1;
287 		}
288 
289 	return 0;
290 }
291 
292 static int dmar_pci_bus_add_dev(struct dmar_pci_notify_info *info)
293 {
294 	int ret = 0;
295 	struct dmar_drhd_unit *dmaru;
296 	struct acpi_dmar_hardware_unit *drhd;
297 
298 	for_each_drhd_unit(dmaru) {
299 		if (dmaru->include_all)
300 			continue;
301 
302 		drhd = container_of(dmaru->hdr,
303 				    struct acpi_dmar_hardware_unit, header);
304 		ret = dmar_insert_dev_scope(info, (void *)(drhd + 1),
305 				((void *)drhd) + drhd->header.length,
306 				dmaru->segment,
307 				dmaru->devices, dmaru->devices_cnt);
308 		if (ret)
309 			break;
310 	}
311 	if (ret >= 0)
312 		ret = dmar_iommu_notify_scope_dev(info);
313 	if (ret < 0 && dmar_dev_scope_status == 0)
314 		dmar_dev_scope_status = ret;
315 
316 	if (ret >= 0)
317 		intel_irq_remap_add_device(info);
318 
319 	return ret;
320 }
321 
322 static void  dmar_pci_bus_del_dev(struct dmar_pci_notify_info *info)
323 {
324 	struct dmar_drhd_unit *dmaru;
325 
326 	for_each_drhd_unit(dmaru)
327 		if (dmar_remove_dev_scope(info, dmaru->segment,
328 			dmaru->devices, dmaru->devices_cnt))
329 			break;
330 	dmar_iommu_notify_scope_dev(info);
331 }
332 
333 static inline void vf_inherit_msi_domain(struct pci_dev *pdev)
334 {
335 	struct pci_dev *physfn = pci_physfn(pdev);
336 
337 	dev_set_msi_domain(&pdev->dev, dev_get_msi_domain(&physfn->dev));
338 }
339 
340 static int dmar_pci_bus_notifier(struct notifier_block *nb,
341 				 unsigned long action, void *data)
342 {
343 	struct pci_dev *pdev = to_pci_dev(data);
344 	struct dmar_pci_notify_info *info;
345 
346 	/* Only care about add/remove events for physical functions.
347 	 * For VFs we actually do the lookup based on the corresponding
348 	 * PF in device_to_iommu() anyway. */
349 	if (pdev->is_virtfn) {
350 		/*
351 		 * Ensure that the VF device inherits the irq domain of the
352 		 * PF device. Ideally the device would inherit the domain
353 		 * from the bus, but DMAR can have multiple units per bus
354 		 * which makes this impossible. The VF 'bus' could inherit
355 		 * from the PF device, but that's yet another x86'sism to
356 		 * inflict on everybody else.
357 		 */
358 		if (action == BUS_NOTIFY_ADD_DEVICE)
359 			vf_inherit_msi_domain(pdev);
360 		return NOTIFY_DONE;
361 	}
362 
363 	if (action != BUS_NOTIFY_ADD_DEVICE &&
364 	    action != BUS_NOTIFY_REMOVED_DEVICE)
365 		return NOTIFY_DONE;
366 
367 	info = dmar_alloc_pci_notify_info(pdev, action);
368 	if (!info)
369 		return NOTIFY_DONE;
370 
371 	down_write(&dmar_global_lock);
372 	if (action == BUS_NOTIFY_ADD_DEVICE)
373 		dmar_pci_bus_add_dev(info);
374 	else if (action == BUS_NOTIFY_REMOVED_DEVICE)
375 		dmar_pci_bus_del_dev(info);
376 	up_write(&dmar_global_lock);
377 
378 	dmar_free_pci_notify_info(info);
379 
380 	return NOTIFY_OK;
381 }
382 
383 static struct notifier_block dmar_pci_bus_nb = {
384 	.notifier_call = dmar_pci_bus_notifier,
385 	.priority = 1,
386 };
387 
388 static struct dmar_drhd_unit *
389 dmar_find_dmaru(struct acpi_dmar_hardware_unit *drhd)
390 {
391 	struct dmar_drhd_unit *dmaru;
392 
393 	list_for_each_entry_rcu(dmaru, &dmar_drhd_units, list,
394 				dmar_rcu_check())
395 		if (dmaru->segment == drhd->segment &&
396 		    dmaru->reg_base_addr == drhd->address)
397 			return dmaru;
398 
399 	return NULL;
400 }
401 
402 /*
403  * dmar_parse_one_drhd - parses exactly one DMA remapping hardware definition
404  * structure which uniquely represent one DMA remapping hardware unit
405  * present in the platform
406  */
407 static int dmar_parse_one_drhd(struct acpi_dmar_header *header, void *arg)
408 {
409 	struct acpi_dmar_hardware_unit *drhd;
410 	struct dmar_drhd_unit *dmaru;
411 	int ret;
412 
413 	drhd = (struct acpi_dmar_hardware_unit *)header;
414 	dmaru = dmar_find_dmaru(drhd);
415 	if (dmaru)
416 		goto out;
417 
418 	dmaru = kzalloc(sizeof(*dmaru) + header->length, GFP_KERNEL);
419 	if (!dmaru)
420 		return -ENOMEM;
421 
422 	/*
423 	 * If header is allocated from slab by ACPI _DSM method, we need to
424 	 * copy the content because the memory buffer will be freed on return.
425 	 */
426 	dmaru->hdr = (void *)(dmaru + 1);
427 	memcpy(dmaru->hdr, header, header->length);
428 	dmaru->reg_base_addr = drhd->address;
429 	dmaru->segment = drhd->segment;
430 	dmaru->include_all = drhd->flags & 0x1; /* BIT0: INCLUDE_ALL */
431 	dmaru->devices = dmar_alloc_dev_scope((void *)(drhd + 1),
432 					      ((void *)drhd) + drhd->header.length,
433 					      &dmaru->devices_cnt);
434 	if (dmaru->devices_cnt && dmaru->devices == NULL) {
435 		kfree(dmaru);
436 		return -ENOMEM;
437 	}
438 
439 	ret = alloc_iommu(dmaru);
440 	if (ret) {
441 		dmar_free_dev_scope(&dmaru->devices,
442 				    &dmaru->devices_cnt);
443 		kfree(dmaru);
444 		return ret;
445 	}
446 	dmar_register_drhd_unit(dmaru);
447 
448 out:
449 	if (arg)
450 		(*(int *)arg)++;
451 
452 	return 0;
453 }
454 
455 static void dmar_free_drhd(struct dmar_drhd_unit *dmaru)
456 {
457 	if (dmaru->devices && dmaru->devices_cnt)
458 		dmar_free_dev_scope(&dmaru->devices, &dmaru->devices_cnt);
459 	if (dmaru->iommu)
460 		free_iommu(dmaru->iommu);
461 	kfree(dmaru);
462 }
463 
464 static int __init dmar_parse_one_andd(struct acpi_dmar_header *header,
465 				      void *arg)
466 {
467 	struct acpi_dmar_andd *andd = (void *)header;
468 
469 	/* Check for NUL termination within the designated length */
470 	if (strnlen(andd->device_name, header->length - 8) == header->length - 8) {
471 		pr_warn(FW_BUG
472 			   "Your BIOS is broken; ANDD object name is not NUL-terminated\n"
473 			   "BIOS vendor: %s; Ver: %s; Product Version: %s\n",
474 			   dmi_get_system_info(DMI_BIOS_VENDOR),
475 			   dmi_get_system_info(DMI_BIOS_VERSION),
476 			   dmi_get_system_info(DMI_PRODUCT_VERSION));
477 		add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
478 		return -EINVAL;
479 	}
480 	pr_info("ANDD device: %x name: %s\n", andd->device_number,
481 		andd->device_name);
482 
483 	return 0;
484 }
485 
486 #ifdef CONFIG_ACPI_NUMA
487 static int dmar_parse_one_rhsa(struct acpi_dmar_header *header, void *arg)
488 {
489 	struct acpi_dmar_rhsa *rhsa;
490 	struct dmar_drhd_unit *drhd;
491 
492 	rhsa = (struct acpi_dmar_rhsa *)header;
493 	for_each_drhd_unit(drhd) {
494 		if (drhd->reg_base_addr == rhsa->base_address) {
495 			int node = pxm_to_node(rhsa->proximity_domain);
496 
497 			if (node != NUMA_NO_NODE && !node_online(node))
498 				node = NUMA_NO_NODE;
499 			drhd->iommu->node = node;
500 			return 0;
501 		}
502 	}
503 	pr_warn(FW_BUG
504 		"Your BIOS is broken; RHSA refers to non-existent DMAR unit at %llx\n"
505 		"BIOS vendor: %s; Ver: %s; Product Version: %s\n",
506 		rhsa->base_address,
507 		dmi_get_system_info(DMI_BIOS_VENDOR),
508 		dmi_get_system_info(DMI_BIOS_VERSION),
509 		dmi_get_system_info(DMI_PRODUCT_VERSION));
510 	add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
511 
512 	return 0;
513 }
514 #else
515 #define	dmar_parse_one_rhsa		dmar_res_noop
516 #endif
517 
518 static void
519 dmar_table_print_dmar_entry(struct acpi_dmar_header *header)
520 {
521 	struct acpi_dmar_hardware_unit *drhd;
522 	struct acpi_dmar_reserved_memory *rmrr;
523 	struct acpi_dmar_atsr *atsr;
524 	struct acpi_dmar_rhsa *rhsa;
525 	struct acpi_dmar_satc *satc;
526 
527 	switch (header->type) {
528 	case ACPI_DMAR_TYPE_HARDWARE_UNIT:
529 		drhd = container_of(header, struct acpi_dmar_hardware_unit,
530 				    header);
531 		pr_info("DRHD base: %#016Lx flags: %#x\n",
532 			(unsigned long long)drhd->address, drhd->flags);
533 		break;
534 	case ACPI_DMAR_TYPE_RESERVED_MEMORY:
535 		rmrr = container_of(header, struct acpi_dmar_reserved_memory,
536 				    header);
537 		pr_info("RMRR base: %#016Lx end: %#016Lx\n",
538 			(unsigned long long)rmrr->base_address,
539 			(unsigned long long)rmrr->end_address);
540 		break;
541 	case ACPI_DMAR_TYPE_ROOT_ATS:
542 		atsr = container_of(header, struct acpi_dmar_atsr, header);
543 		pr_info("ATSR flags: %#x\n", atsr->flags);
544 		break;
545 	case ACPI_DMAR_TYPE_HARDWARE_AFFINITY:
546 		rhsa = container_of(header, struct acpi_dmar_rhsa, header);
547 		pr_info("RHSA base: %#016Lx proximity domain: %#x\n",
548 		       (unsigned long long)rhsa->base_address,
549 		       rhsa->proximity_domain);
550 		break;
551 	case ACPI_DMAR_TYPE_NAMESPACE:
552 		/* We don't print this here because we need to sanity-check
553 		   it first. So print it in dmar_parse_one_andd() instead. */
554 		break;
555 	case ACPI_DMAR_TYPE_SATC:
556 		satc = container_of(header, struct acpi_dmar_satc, header);
557 		pr_info("SATC flags: 0x%x\n", satc->flags);
558 		break;
559 	}
560 }
561 
562 /**
563  * dmar_table_detect - checks to see if the platform supports DMAR devices
564  */
565 static int __init dmar_table_detect(void)
566 {
567 	acpi_status status = AE_OK;
568 
569 	/* if we could find DMAR table, then there are DMAR devices */
570 	status = acpi_get_table(ACPI_SIG_DMAR, 0, &dmar_tbl);
571 
572 	if (ACPI_SUCCESS(status) && !dmar_tbl) {
573 		pr_warn("Unable to map DMAR\n");
574 		status = AE_NOT_FOUND;
575 	}
576 
577 	return ACPI_SUCCESS(status) ? 0 : -ENOENT;
578 }
579 
580 static int dmar_walk_remapping_entries(struct acpi_dmar_header *start,
581 				       size_t len, struct dmar_res_callback *cb)
582 {
583 	struct acpi_dmar_header *iter, *next;
584 	struct acpi_dmar_header *end = ((void *)start) + len;
585 
586 	for (iter = start; iter < end; iter = next) {
587 		next = (void *)iter + iter->length;
588 		if (iter->length == 0) {
589 			/* Avoid looping forever on bad ACPI tables */
590 			pr_debug(FW_BUG "Invalid 0-length structure\n");
591 			break;
592 		} else if (next > end) {
593 			/* Avoid passing table end */
594 			pr_warn(FW_BUG "Record passes table end\n");
595 			return -EINVAL;
596 		}
597 
598 		if (cb->print_entry)
599 			dmar_table_print_dmar_entry(iter);
600 
601 		if (iter->type >= ACPI_DMAR_TYPE_RESERVED) {
602 			/* continue for forward compatibility */
603 			pr_debug("Unknown DMAR structure type %d\n",
604 				 iter->type);
605 		} else if (cb->cb[iter->type]) {
606 			int ret;
607 
608 			ret = cb->cb[iter->type](iter, cb->arg[iter->type]);
609 			if (ret)
610 				return ret;
611 		} else if (!cb->ignore_unhandled) {
612 			pr_warn("No handler for DMAR structure type %d\n",
613 				iter->type);
614 			return -EINVAL;
615 		}
616 	}
617 
618 	return 0;
619 }
620 
621 static inline int dmar_walk_dmar_table(struct acpi_table_dmar *dmar,
622 				       struct dmar_res_callback *cb)
623 {
624 	return dmar_walk_remapping_entries((void *)(dmar + 1),
625 			dmar->header.length - sizeof(*dmar), cb);
626 }
627 
628 /**
629  * parse_dmar_table - parses the DMA reporting table
630  */
631 static int __init
632 parse_dmar_table(void)
633 {
634 	struct acpi_table_dmar *dmar;
635 	int drhd_count = 0;
636 	int ret;
637 	struct dmar_res_callback cb = {
638 		.print_entry = true,
639 		.ignore_unhandled = true,
640 		.arg[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &drhd_count,
641 		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_parse_one_drhd,
642 		.cb[ACPI_DMAR_TYPE_RESERVED_MEMORY] = &dmar_parse_one_rmrr,
643 		.cb[ACPI_DMAR_TYPE_ROOT_ATS] = &dmar_parse_one_atsr,
644 		.cb[ACPI_DMAR_TYPE_HARDWARE_AFFINITY] = &dmar_parse_one_rhsa,
645 		.cb[ACPI_DMAR_TYPE_NAMESPACE] = &dmar_parse_one_andd,
646 		.cb[ACPI_DMAR_TYPE_SATC] = &dmar_parse_one_satc,
647 	};
648 
649 	/*
650 	 * Do it again, earlier dmar_tbl mapping could be mapped with
651 	 * fixed map.
652 	 */
653 	dmar_table_detect();
654 
655 	/*
656 	 * ACPI tables may not be DMA protected by tboot, so use DMAR copy
657 	 * SINIT saved in SinitMleData in TXT heap (which is DMA protected)
658 	 */
659 	dmar_tbl = tboot_get_dmar_table(dmar_tbl);
660 
661 	dmar = (struct acpi_table_dmar *)dmar_tbl;
662 	if (!dmar)
663 		return -ENODEV;
664 
665 	if (dmar->width < PAGE_SHIFT - 1) {
666 		pr_warn("Invalid DMAR haw\n");
667 		return -EINVAL;
668 	}
669 
670 	pr_info("Host address width %d\n", dmar->width + 1);
671 	ret = dmar_walk_dmar_table(dmar, &cb);
672 	if (ret == 0 && drhd_count == 0)
673 		pr_warn(FW_BUG "No DRHD structure found in DMAR table\n");
674 
675 	return ret;
676 }
677 
678 static int dmar_pci_device_match(struct dmar_dev_scope devices[],
679 				 int cnt, struct pci_dev *dev)
680 {
681 	int index;
682 	struct device *tmp;
683 
684 	while (dev) {
685 		for_each_active_dev_scope(devices, cnt, index, tmp)
686 			if (dev_is_pci(tmp) && dev == to_pci_dev(tmp))
687 				return 1;
688 
689 		/* Check our parent */
690 		dev = dev->bus->self;
691 	}
692 
693 	return 0;
694 }
695 
696 struct dmar_drhd_unit *
697 dmar_find_matched_drhd_unit(struct pci_dev *dev)
698 {
699 	struct dmar_drhd_unit *dmaru;
700 	struct acpi_dmar_hardware_unit *drhd;
701 
702 	dev = pci_physfn(dev);
703 
704 	rcu_read_lock();
705 	for_each_drhd_unit(dmaru) {
706 		drhd = container_of(dmaru->hdr,
707 				    struct acpi_dmar_hardware_unit,
708 				    header);
709 
710 		if (dmaru->include_all &&
711 		    drhd->segment == pci_domain_nr(dev->bus))
712 			goto out;
713 
714 		if (dmar_pci_device_match(dmaru->devices,
715 					  dmaru->devices_cnt, dev))
716 			goto out;
717 	}
718 	dmaru = NULL;
719 out:
720 	rcu_read_unlock();
721 
722 	return dmaru;
723 }
724 
725 static void __init dmar_acpi_insert_dev_scope(u8 device_number,
726 					      struct acpi_device *adev)
727 {
728 	struct dmar_drhd_unit *dmaru;
729 	struct acpi_dmar_hardware_unit *drhd;
730 	struct acpi_dmar_device_scope *scope;
731 	struct device *tmp;
732 	int i;
733 	struct acpi_dmar_pci_path *path;
734 
735 	for_each_drhd_unit(dmaru) {
736 		drhd = container_of(dmaru->hdr,
737 				    struct acpi_dmar_hardware_unit,
738 				    header);
739 
740 		for (scope = (void *)(drhd + 1);
741 		     (unsigned long)scope < ((unsigned long)drhd) + drhd->header.length;
742 		     scope = ((void *)scope) + scope->length) {
743 			if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_NAMESPACE)
744 				continue;
745 			if (scope->enumeration_id != device_number)
746 				continue;
747 
748 			path = (void *)(scope + 1);
749 			pr_info("ACPI device \"%s\" under DMAR at %llx as %02x:%02x.%d\n",
750 				dev_name(&adev->dev), dmaru->reg_base_addr,
751 				scope->bus, path->device, path->function);
752 			for_each_dev_scope(dmaru->devices, dmaru->devices_cnt, i, tmp)
753 				if (tmp == NULL) {
754 					dmaru->devices[i].bus = scope->bus;
755 					dmaru->devices[i].devfn = PCI_DEVFN(path->device,
756 									    path->function);
757 					rcu_assign_pointer(dmaru->devices[i].dev,
758 							   get_device(&adev->dev));
759 					return;
760 				}
761 			BUG_ON(i >= dmaru->devices_cnt);
762 		}
763 	}
764 	pr_warn("No IOMMU scope found for ANDD enumeration ID %d (%s)\n",
765 		device_number, dev_name(&adev->dev));
766 }
767 
768 static int __init dmar_acpi_dev_scope_init(void)
769 {
770 	struct acpi_dmar_andd *andd;
771 
772 	if (dmar_tbl == NULL)
773 		return -ENODEV;
774 
775 	for (andd = (void *)dmar_tbl + sizeof(struct acpi_table_dmar);
776 	     ((unsigned long)andd) < ((unsigned long)dmar_tbl) + dmar_tbl->length;
777 	     andd = ((void *)andd) + andd->header.length) {
778 		if (andd->header.type == ACPI_DMAR_TYPE_NAMESPACE) {
779 			acpi_handle h;
780 			struct acpi_device *adev;
781 
782 			if (!ACPI_SUCCESS(acpi_get_handle(ACPI_ROOT_OBJECT,
783 							  andd->device_name,
784 							  &h))) {
785 				pr_err("Failed to find handle for ACPI object %s\n",
786 				       andd->device_name);
787 				continue;
788 			}
789 			adev = acpi_fetch_acpi_dev(h);
790 			if (!adev) {
791 				pr_err("Failed to get device for ACPI object %s\n",
792 				       andd->device_name);
793 				continue;
794 			}
795 			dmar_acpi_insert_dev_scope(andd->device_number, adev);
796 		}
797 	}
798 	return 0;
799 }
800 
801 int __init dmar_dev_scope_init(void)
802 {
803 	struct pci_dev *dev = NULL;
804 	struct dmar_pci_notify_info *info;
805 
806 	if (dmar_dev_scope_status != 1)
807 		return dmar_dev_scope_status;
808 
809 	if (list_empty(&dmar_drhd_units)) {
810 		dmar_dev_scope_status = -ENODEV;
811 	} else {
812 		dmar_dev_scope_status = 0;
813 
814 		dmar_acpi_dev_scope_init();
815 
816 		for_each_pci_dev(dev) {
817 			if (dev->is_virtfn)
818 				continue;
819 
820 			info = dmar_alloc_pci_notify_info(dev,
821 					BUS_NOTIFY_ADD_DEVICE);
822 			if (!info) {
823 				return dmar_dev_scope_status;
824 			} else {
825 				dmar_pci_bus_add_dev(info);
826 				dmar_free_pci_notify_info(info);
827 			}
828 		}
829 	}
830 
831 	return dmar_dev_scope_status;
832 }
833 
834 void __init dmar_register_bus_notifier(void)
835 {
836 	bus_register_notifier(&pci_bus_type, &dmar_pci_bus_nb);
837 }
838 
839 
840 int __init dmar_table_init(void)
841 {
842 	static int dmar_table_initialized;
843 	int ret;
844 
845 	if (dmar_table_initialized == 0) {
846 		ret = parse_dmar_table();
847 		if (ret < 0) {
848 			if (ret != -ENODEV)
849 				pr_info("Parse DMAR table failure.\n");
850 		} else  if (list_empty(&dmar_drhd_units)) {
851 			pr_info("No DMAR devices found\n");
852 			ret = -ENODEV;
853 		}
854 
855 		if (ret < 0)
856 			dmar_table_initialized = ret;
857 		else
858 			dmar_table_initialized = 1;
859 	}
860 
861 	return dmar_table_initialized < 0 ? dmar_table_initialized : 0;
862 }
863 
864 static void warn_invalid_dmar(u64 addr, const char *message)
865 {
866 	pr_warn_once(FW_BUG
867 		"Your BIOS is broken; DMAR reported at address %llx%s!\n"
868 		"BIOS vendor: %s; Ver: %s; Product Version: %s\n",
869 		addr, message,
870 		dmi_get_system_info(DMI_BIOS_VENDOR),
871 		dmi_get_system_info(DMI_BIOS_VERSION),
872 		dmi_get_system_info(DMI_PRODUCT_VERSION));
873 	add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
874 }
875 
876 static int __ref
877 dmar_validate_one_drhd(struct acpi_dmar_header *entry, void *arg)
878 {
879 	struct acpi_dmar_hardware_unit *drhd;
880 	void __iomem *addr;
881 	u64 cap, ecap;
882 
883 	drhd = (void *)entry;
884 	if (!drhd->address) {
885 		warn_invalid_dmar(0, "");
886 		return -EINVAL;
887 	}
888 
889 	if (arg)
890 		addr = ioremap(drhd->address, VTD_PAGE_SIZE);
891 	else
892 		addr = early_ioremap(drhd->address, VTD_PAGE_SIZE);
893 	if (!addr) {
894 		pr_warn("Can't validate DRHD address: %llx\n", drhd->address);
895 		return -EINVAL;
896 	}
897 
898 	cap = dmar_readq(addr + DMAR_CAP_REG);
899 	ecap = dmar_readq(addr + DMAR_ECAP_REG);
900 
901 	if (arg)
902 		iounmap(addr);
903 	else
904 		early_iounmap(addr, VTD_PAGE_SIZE);
905 
906 	if (cap == (uint64_t)-1 && ecap == (uint64_t)-1) {
907 		warn_invalid_dmar(drhd->address, " returns all ones");
908 		return -EINVAL;
909 	}
910 
911 	return 0;
912 }
913 
914 void __init detect_intel_iommu(void)
915 {
916 	int ret;
917 	struct dmar_res_callback validate_drhd_cb = {
918 		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_validate_one_drhd,
919 		.ignore_unhandled = true,
920 	};
921 
922 	down_write(&dmar_global_lock);
923 	ret = dmar_table_detect();
924 	if (!ret)
925 		ret = dmar_walk_dmar_table((struct acpi_table_dmar *)dmar_tbl,
926 					   &validate_drhd_cb);
927 	if (!ret && !no_iommu && !iommu_detected &&
928 	    (!dmar_disabled || dmar_platform_optin())) {
929 		iommu_detected = 1;
930 		/* Make sure ACS will be enabled */
931 		pci_request_acs();
932 	}
933 
934 #ifdef CONFIG_X86
935 	if (!ret) {
936 		x86_init.iommu.iommu_init = intel_iommu_init;
937 		x86_platform.iommu_shutdown = intel_iommu_shutdown;
938 	}
939 
940 #endif
941 
942 	if (dmar_tbl) {
943 		acpi_put_table(dmar_tbl);
944 		dmar_tbl = NULL;
945 	}
946 	up_write(&dmar_global_lock);
947 }
948 
949 static void unmap_iommu(struct intel_iommu *iommu)
950 {
951 	iounmap(iommu->reg);
952 	release_mem_region(iommu->reg_phys, iommu->reg_size);
953 }
954 
955 /**
956  * map_iommu: map the iommu's registers
957  * @iommu: the iommu to map
958  * @phys_addr: the physical address of the base resgister
959  *
960  * Memory map the iommu's registers.  Start w/ a single page, and
961  * possibly expand if that turns out to be insufficent.
962  */
963 static int map_iommu(struct intel_iommu *iommu, u64 phys_addr)
964 {
965 	int map_size, err=0;
966 
967 	iommu->reg_phys = phys_addr;
968 	iommu->reg_size = VTD_PAGE_SIZE;
969 
970 	if (!request_mem_region(iommu->reg_phys, iommu->reg_size, iommu->name)) {
971 		pr_err("Can't reserve memory\n");
972 		err = -EBUSY;
973 		goto out;
974 	}
975 
976 	iommu->reg = ioremap(iommu->reg_phys, iommu->reg_size);
977 	if (!iommu->reg) {
978 		pr_err("Can't map the region\n");
979 		err = -ENOMEM;
980 		goto release;
981 	}
982 
983 	iommu->cap = dmar_readq(iommu->reg + DMAR_CAP_REG);
984 	iommu->ecap = dmar_readq(iommu->reg + DMAR_ECAP_REG);
985 
986 	if (iommu->cap == (uint64_t)-1 && iommu->ecap == (uint64_t)-1) {
987 		err = -EINVAL;
988 		warn_invalid_dmar(phys_addr, " returns all ones");
989 		goto unmap;
990 	}
991 	if (ecap_vcs(iommu->ecap))
992 		iommu->vccap = dmar_readq(iommu->reg + DMAR_VCCAP_REG);
993 
994 	/* the registers might be more than one page */
995 	map_size = max_t(int, ecap_max_iotlb_offset(iommu->ecap),
996 			 cap_max_fault_reg_offset(iommu->cap));
997 	map_size = VTD_PAGE_ALIGN(map_size);
998 	if (map_size > iommu->reg_size) {
999 		iounmap(iommu->reg);
1000 		release_mem_region(iommu->reg_phys, iommu->reg_size);
1001 		iommu->reg_size = map_size;
1002 		if (!request_mem_region(iommu->reg_phys, iommu->reg_size,
1003 					iommu->name)) {
1004 			pr_err("Can't reserve memory\n");
1005 			err = -EBUSY;
1006 			goto out;
1007 		}
1008 		iommu->reg = ioremap(iommu->reg_phys, iommu->reg_size);
1009 		if (!iommu->reg) {
1010 			pr_err("Can't map the region\n");
1011 			err = -ENOMEM;
1012 			goto release;
1013 		}
1014 	}
1015 	err = 0;
1016 	goto out;
1017 
1018 unmap:
1019 	iounmap(iommu->reg);
1020 release:
1021 	release_mem_region(iommu->reg_phys, iommu->reg_size);
1022 out:
1023 	return err;
1024 }
1025 
1026 static int alloc_iommu(struct dmar_drhd_unit *drhd)
1027 {
1028 	struct intel_iommu *iommu;
1029 	u32 ver, sts;
1030 	int agaw = -1;
1031 	int msagaw = -1;
1032 	int err;
1033 
1034 	if (!drhd->reg_base_addr) {
1035 		warn_invalid_dmar(0, "");
1036 		return -EINVAL;
1037 	}
1038 
1039 	iommu = kzalloc(sizeof(*iommu), GFP_KERNEL);
1040 	if (!iommu)
1041 		return -ENOMEM;
1042 
1043 	iommu->seq_id = ida_alloc_range(&dmar_seq_ids, 0,
1044 					DMAR_UNITS_SUPPORTED - 1, GFP_KERNEL);
1045 	if (iommu->seq_id < 0) {
1046 		pr_err("Failed to allocate seq_id\n");
1047 		err = iommu->seq_id;
1048 		goto error;
1049 	}
1050 	sprintf(iommu->name, "dmar%d", iommu->seq_id);
1051 
1052 	err = map_iommu(iommu, drhd->reg_base_addr);
1053 	if (err) {
1054 		pr_err("Failed to map %s\n", iommu->name);
1055 		goto error_free_seq_id;
1056 	}
1057 
1058 	err = -EINVAL;
1059 	if (cap_sagaw(iommu->cap) == 0) {
1060 		pr_info("%s: No supported address widths. Not attempting DMA translation.\n",
1061 			iommu->name);
1062 		drhd->ignored = 1;
1063 	}
1064 
1065 	if (!drhd->ignored) {
1066 		agaw = iommu_calculate_agaw(iommu);
1067 		if (agaw < 0) {
1068 			pr_err("Cannot get a valid agaw for iommu (seq_id = %d)\n",
1069 			       iommu->seq_id);
1070 			drhd->ignored = 1;
1071 		}
1072 	}
1073 	if (!drhd->ignored) {
1074 		msagaw = iommu_calculate_max_sagaw(iommu);
1075 		if (msagaw < 0) {
1076 			pr_err("Cannot get a valid max agaw for iommu (seq_id = %d)\n",
1077 			       iommu->seq_id);
1078 			drhd->ignored = 1;
1079 			agaw = -1;
1080 		}
1081 	}
1082 	iommu->agaw = agaw;
1083 	iommu->msagaw = msagaw;
1084 	iommu->segment = drhd->segment;
1085 
1086 	iommu->node = NUMA_NO_NODE;
1087 
1088 	ver = readl(iommu->reg + DMAR_VER_REG);
1089 	pr_info("%s: reg_base_addr %llx ver %d:%d cap %llx ecap %llx\n",
1090 		iommu->name,
1091 		(unsigned long long)drhd->reg_base_addr,
1092 		DMAR_VER_MAJOR(ver), DMAR_VER_MINOR(ver),
1093 		(unsigned long long)iommu->cap,
1094 		(unsigned long long)iommu->ecap);
1095 
1096 	/* Reflect status in gcmd */
1097 	sts = readl(iommu->reg + DMAR_GSTS_REG);
1098 	if (sts & DMA_GSTS_IRES)
1099 		iommu->gcmd |= DMA_GCMD_IRE;
1100 	if (sts & DMA_GSTS_TES)
1101 		iommu->gcmd |= DMA_GCMD_TE;
1102 	if (sts & DMA_GSTS_QIES)
1103 		iommu->gcmd |= DMA_GCMD_QIE;
1104 
1105 	raw_spin_lock_init(&iommu->register_lock);
1106 
1107 	/*
1108 	 * This is only for hotplug; at boot time intel_iommu_enabled won't
1109 	 * be set yet. When intel_iommu_init() runs, it registers the units
1110 	 * present at boot time, then sets intel_iommu_enabled.
1111 	 */
1112 	if (intel_iommu_enabled && !drhd->ignored) {
1113 		err = iommu_device_sysfs_add(&iommu->iommu, NULL,
1114 					     intel_iommu_groups,
1115 					     "%s", iommu->name);
1116 		if (err)
1117 			goto err_unmap;
1118 
1119 		err = iommu_device_register(&iommu->iommu, &intel_iommu_ops, NULL);
1120 		if (err)
1121 			goto err_sysfs;
1122 	}
1123 
1124 	drhd->iommu = iommu;
1125 	iommu->drhd = drhd;
1126 
1127 	return 0;
1128 
1129 err_sysfs:
1130 	iommu_device_sysfs_remove(&iommu->iommu);
1131 err_unmap:
1132 	unmap_iommu(iommu);
1133 error_free_seq_id:
1134 	ida_free(&dmar_seq_ids, iommu->seq_id);
1135 error:
1136 	kfree(iommu);
1137 	return err;
1138 }
1139 
1140 static void free_iommu(struct intel_iommu *iommu)
1141 {
1142 	if (intel_iommu_enabled && !iommu->drhd->ignored) {
1143 		iommu_device_unregister(&iommu->iommu);
1144 		iommu_device_sysfs_remove(&iommu->iommu);
1145 	}
1146 
1147 	if (iommu->irq) {
1148 		if (iommu->pr_irq) {
1149 			free_irq(iommu->pr_irq, iommu);
1150 			dmar_free_hwirq(iommu->pr_irq);
1151 			iommu->pr_irq = 0;
1152 		}
1153 		free_irq(iommu->irq, iommu);
1154 		dmar_free_hwirq(iommu->irq);
1155 		iommu->irq = 0;
1156 	}
1157 
1158 	if (iommu->qi) {
1159 		free_page((unsigned long)iommu->qi->desc);
1160 		kfree(iommu->qi->desc_status);
1161 		kfree(iommu->qi);
1162 	}
1163 
1164 	if (iommu->reg)
1165 		unmap_iommu(iommu);
1166 
1167 	ida_free(&dmar_seq_ids, iommu->seq_id);
1168 	kfree(iommu);
1169 }
1170 
1171 /*
1172  * Reclaim all the submitted descriptors which have completed its work.
1173  */
1174 static inline void reclaim_free_desc(struct q_inval *qi)
1175 {
1176 	while (qi->desc_status[qi->free_tail] == QI_DONE ||
1177 	       qi->desc_status[qi->free_tail] == QI_ABORT) {
1178 		qi->desc_status[qi->free_tail] = QI_FREE;
1179 		qi->free_tail = (qi->free_tail + 1) % QI_LENGTH;
1180 		qi->free_cnt++;
1181 	}
1182 }
1183 
1184 static const char *qi_type_string(u8 type)
1185 {
1186 	switch (type) {
1187 	case QI_CC_TYPE:
1188 		return "Context-cache Invalidation";
1189 	case QI_IOTLB_TYPE:
1190 		return "IOTLB Invalidation";
1191 	case QI_DIOTLB_TYPE:
1192 		return "Device-TLB Invalidation";
1193 	case QI_IEC_TYPE:
1194 		return "Interrupt Entry Cache Invalidation";
1195 	case QI_IWD_TYPE:
1196 		return "Invalidation Wait";
1197 	case QI_EIOTLB_TYPE:
1198 		return "PASID-based IOTLB Invalidation";
1199 	case QI_PC_TYPE:
1200 		return "PASID-cache Invalidation";
1201 	case QI_DEIOTLB_TYPE:
1202 		return "PASID-based Device-TLB Invalidation";
1203 	case QI_PGRP_RESP_TYPE:
1204 		return "Page Group Response";
1205 	default:
1206 		return "UNKNOWN";
1207 	}
1208 }
1209 
1210 static void qi_dump_fault(struct intel_iommu *iommu, u32 fault)
1211 {
1212 	unsigned int head = dmar_readl(iommu->reg + DMAR_IQH_REG);
1213 	u64 iqe_err = dmar_readq(iommu->reg + DMAR_IQER_REG);
1214 	struct qi_desc *desc = iommu->qi->desc + head;
1215 
1216 	if (fault & DMA_FSTS_IQE)
1217 		pr_err("VT-d detected Invalidation Queue Error: Reason %llx",
1218 		       DMAR_IQER_REG_IQEI(iqe_err));
1219 	if (fault & DMA_FSTS_ITE)
1220 		pr_err("VT-d detected Invalidation Time-out Error: SID %llx",
1221 		       DMAR_IQER_REG_ITESID(iqe_err));
1222 	if (fault & DMA_FSTS_ICE)
1223 		pr_err("VT-d detected Invalidation Completion Error: SID %llx",
1224 		       DMAR_IQER_REG_ICESID(iqe_err));
1225 
1226 	pr_err("QI HEAD: %s qw0 = 0x%llx, qw1 = 0x%llx\n",
1227 	       qi_type_string(desc->qw0 & 0xf),
1228 	       (unsigned long long)desc->qw0,
1229 	       (unsigned long long)desc->qw1);
1230 
1231 	head = ((head >> qi_shift(iommu)) + QI_LENGTH - 1) % QI_LENGTH;
1232 	head <<= qi_shift(iommu);
1233 	desc = iommu->qi->desc + head;
1234 
1235 	pr_err("QI PRIOR: %s qw0 = 0x%llx, qw1 = 0x%llx\n",
1236 	       qi_type_string(desc->qw0 & 0xf),
1237 	       (unsigned long long)desc->qw0,
1238 	       (unsigned long long)desc->qw1);
1239 }
1240 
1241 static int qi_check_fault(struct intel_iommu *iommu, int index, int wait_index)
1242 {
1243 	u32 fault;
1244 	int head, tail;
1245 	struct q_inval *qi = iommu->qi;
1246 	int shift = qi_shift(iommu);
1247 
1248 	if (qi->desc_status[wait_index] == QI_ABORT)
1249 		return -EAGAIN;
1250 
1251 	fault = readl(iommu->reg + DMAR_FSTS_REG);
1252 	if (fault & (DMA_FSTS_IQE | DMA_FSTS_ITE | DMA_FSTS_ICE))
1253 		qi_dump_fault(iommu, fault);
1254 
1255 	/*
1256 	 * If IQE happens, the head points to the descriptor associated
1257 	 * with the error. No new descriptors are fetched until the IQE
1258 	 * is cleared.
1259 	 */
1260 	if (fault & DMA_FSTS_IQE) {
1261 		head = readl(iommu->reg + DMAR_IQH_REG);
1262 		if ((head >> shift) == index) {
1263 			struct qi_desc *desc = qi->desc + head;
1264 
1265 			/*
1266 			 * desc->qw2 and desc->qw3 are either reserved or
1267 			 * used by software as private data. We won't print
1268 			 * out these two qw's for security consideration.
1269 			 */
1270 			memcpy(desc, qi->desc + (wait_index << shift),
1271 			       1 << shift);
1272 			writel(DMA_FSTS_IQE, iommu->reg + DMAR_FSTS_REG);
1273 			pr_info("Invalidation Queue Error (IQE) cleared\n");
1274 			return -EINVAL;
1275 		}
1276 	}
1277 
1278 	/*
1279 	 * If ITE happens, all pending wait_desc commands are aborted.
1280 	 * No new descriptors are fetched until the ITE is cleared.
1281 	 */
1282 	if (fault & DMA_FSTS_ITE) {
1283 		head = readl(iommu->reg + DMAR_IQH_REG);
1284 		head = ((head >> shift) - 1 + QI_LENGTH) % QI_LENGTH;
1285 		head |= 1;
1286 		tail = readl(iommu->reg + DMAR_IQT_REG);
1287 		tail = ((tail >> shift) - 1 + QI_LENGTH) % QI_LENGTH;
1288 
1289 		writel(DMA_FSTS_ITE, iommu->reg + DMAR_FSTS_REG);
1290 		pr_info("Invalidation Time-out Error (ITE) cleared\n");
1291 
1292 		do {
1293 			if (qi->desc_status[head] == QI_IN_USE)
1294 				qi->desc_status[head] = QI_ABORT;
1295 			head = (head - 2 + QI_LENGTH) % QI_LENGTH;
1296 		} while (head != tail);
1297 
1298 		if (qi->desc_status[wait_index] == QI_ABORT)
1299 			return -EAGAIN;
1300 	}
1301 
1302 	if (fault & DMA_FSTS_ICE) {
1303 		writel(DMA_FSTS_ICE, iommu->reg + DMAR_FSTS_REG);
1304 		pr_info("Invalidation Completion Error (ICE) cleared\n");
1305 	}
1306 
1307 	return 0;
1308 }
1309 
1310 /*
1311  * Function to submit invalidation descriptors of all types to the queued
1312  * invalidation interface(QI). Multiple descriptors can be submitted at a
1313  * time, a wait descriptor will be appended to each submission to ensure
1314  * hardware has completed the invalidation before return. Wait descriptors
1315  * can be part of the submission but it will not be polled for completion.
1316  */
1317 int qi_submit_sync(struct intel_iommu *iommu, struct qi_desc *desc,
1318 		   unsigned int count, unsigned long options)
1319 {
1320 	struct q_inval *qi = iommu->qi;
1321 	s64 devtlb_start_ktime = 0;
1322 	s64 iotlb_start_ktime = 0;
1323 	s64 iec_start_ktime = 0;
1324 	struct qi_desc wait_desc;
1325 	int wait_index, index;
1326 	unsigned long flags;
1327 	int offset, shift;
1328 	int rc, i;
1329 	u64 type;
1330 
1331 	if (!qi)
1332 		return 0;
1333 
1334 	type = desc->qw0 & GENMASK_ULL(3, 0);
1335 
1336 	if ((type == QI_IOTLB_TYPE || type == QI_EIOTLB_TYPE) &&
1337 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_IOTLB))
1338 		iotlb_start_ktime = ktime_to_ns(ktime_get());
1339 
1340 	if ((type == QI_DIOTLB_TYPE || type == QI_DEIOTLB_TYPE) &&
1341 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_DEVTLB))
1342 		devtlb_start_ktime = ktime_to_ns(ktime_get());
1343 
1344 	if (type == QI_IEC_TYPE &&
1345 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_IEC))
1346 		iec_start_ktime = ktime_to_ns(ktime_get());
1347 
1348 restart:
1349 	rc = 0;
1350 
1351 	raw_spin_lock_irqsave(&qi->q_lock, flags);
1352 	/*
1353 	 * Check if we have enough empty slots in the queue to submit,
1354 	 * the calculation is based on:
1355 	 * # of desc + 1 wait desc + 1 space between head and tail
1356 	 */
1357 	while (qi->free_cnt < count + 2) {
1358 		raw_spin_unlock_irqrestore(&qi->q_lock, flags);
1359 		cpu_relax();
1360 		raw_spin_lock_irqsave(&qi->q_lock, flags);
1361 	}
1362 
1363 	index = qi->free_head;
1364 	wait_index = (index + count) % QI_LENGTH;
1365 	shift = qi_shift(iommu);
1366 
1367 	for (i = 0; i < count; i++) {
1368 		offset = ((index + i) % QI_LENGTH) << shift;
1369 		memcpy(qi->desc + offset, &desc[i], 1 << shift);
1370 		qi->desc_status[(index + i) % QI_LENGTH] = QI_IN_USE;
1371 		trace_qi_submit(iommu, desc[i].qw0, desc[i].qw1,
1372 				desc[i].qw2, desc[i].qw3);
1373 	}
1374 	qi->desc_status[wait_index] = QI_IN_USE;
1375 
1376 	wait_desc.qw0 = QI_IWD_STATUS_DATA(QI_DONE) |
1377 			QI_IWD_STATUS_WRITE | QI_IWD_TYPE;
1378 	if (options & QI_OPT_WAIT_DRAIN)
1379 		wait_desc.qw0 |= QI_IWD_PRQ_DRAIN;
1380 	wait_desc.qw1 = virt_to_phys(&qi->desc_status[wait_index]);
1381 	wait_desc.qw2 = 0;
1382 	wait_desc.qw3 = 0;
1383 
1384 	offset = wait_index << shift;
1385 	memcpy(qi->desc + offset, &wait_desc, 1 << shift);
1386 
1387 	qi->free_head = (qi->free_head + count + 1) % QI_LENGTH;
1388 	qi->free_cnt -= count + 1;
1389 
1390 	/*
1391 	 * update the HW tail register indicating the presence of
1392 	 * new descriptors.
1393 	 */
1394 	writel(qi->free_head << shift, iommu->reg + DMAR_IQT_REG);
1395 
1396 	while (qi->desc_status[wait_index] != QI_DONE) {
1397 		/*
1398 		 * We will leave the interrupts disabled, to prevent interrupt
1399 		 * context to queue another cmd while a cmd is already submitted
1400 		 * and waiting for completion on this cpu. This is to avoid
1401 		 * a deadlock where the interrupt context can wait indefinitely
1402 		 * for free slots in the queue.
1403 		 */
1404 		rc = qi_check_fault(iommu, index, wait_index);
1405 		if (rc)
1406 			break;
1407 
1408 		raw_spin_unlock(&qi->q_lock);
1409 		cpu_relax();
1410 		raw_spin_lock(&qi->q_lock);
1411 	}
1412 
1413 	for (i = 0; i < count; i++)
1414 		qi->desc_status[(index + i) % QI_LENGTH] = QI_DONE;
1415 
1416 	reclaim_free_desc(qi);
1417 	raw_spin_unlock_irqrestore(&qi->q_lock, flags);
1418 
1419 	if (rc == -EAGAIN)
1420 		goto restart;
1421 
1422 	if (iotlb_start_ktime)
1423 		dmar_latency_update(iommu, DMAR_LATENCY_INV_IOTLB,
1424 				ktime_to_ns(ktime_get()) - iotlb_start_ktime);
1425 
1426 	if (devtlb_start_ktime)
1427 		dmar_latency_update(iommu, DMAR_LATENCY_INV_DEVTLB,
1428 				ktime_to_ns(ktime_get()) - devtlb_start_ktime);
1429 
1430 	if (iec_start_ktime)
1431 		dmar_latency_update(iommu, DMAR_LATENCY_INV_IEC,
1432 				ktime_to_ns(ktime_get()) - iec_start_ktime);
1433 
1434 	return rc;
1435 }
1436 
1437 /*
1438  * Flush the global interrupt entry cache.
1439  */
1440 void qi_global_iec(struct intel_iommu *iommu)
1441 {
1442 	struct qi_desc desc;
1443 
1444 	desc.qw0 = QI_IEC_TYPE;
1445 	desc.qw1 = 0;
1446 	desc.qw2 = 0;
1447 	desc.qw3 = 0;
1448 
1449 	/* should never fail */
1450 	qi_submit_sync(iommu, &desc, 1, 0);
1451 }
1452 
1453 void qi_flush_context(struct intel_iommu *iommu, u16 did, u16 sid, u8 fm,
1454 		      u64 type)
1455 {
1456 	struct qi_desc desc;
1457 
1458 	desc.qw0 = QI_CC_FM(fm) | QI_CC_SID(sid) | QI_CC_DID(did)
1459 			| QI_CC_GRAN(type) | QI_CC_TYPE;
1460 	desc.qw1 = 0;
1461 	desc.qw2 = 0;
1462 	desc.qw3 = 0;
1463 
1464 	qi_submit_sync(iommu, &desc, 1, 0);
1465 }
1466 
1467 void qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr,
1468 		    unsigned int size_order, u64 type)
1469 {
1470 	u8 dw = 0, dr = 0;
1471 
1472 	struct qi_desc desc;
1473 	int ih = 0;
1474 
1475 	if (cap_write_drain(iommu->cap))
1476 		dw = 1;
1477 
1478 	if (cap_read_drain(iommu->cap))
1479 		dr = 1;
1480 
1481 	desc.qw0 = QI_IOTLB_DID(did) | QI_IOTLB_DR(dr) | QI_IOTLB_DW(dw)
1482 		| QI_IOTLB_GRAN(type) | QI_IOTLB_TYPE;
1483 	desc.qw1 = QI_IOTLB_ADDR(addr) | QI_IOTLB_IH(ih)
1484 		| QI_IOTLB_AM(size_order);
1485 	desc.qw2 = 0;
1486 	desc.qw3 = 0;
1487 
1488 	qi_submit_sync(iommu, &desc, 1, 0);
1489 }
1490 
1491 void qi_flush_dev_iotlb(struct intel_iommu *iommu, u16 sid, u16 pfsid,
1492 			u16 qdep, u64 addr, unsigned mask)
1493 {
1494 	struct qi_desc desc;
1495 
1496 	if (mask) {
1497 		addr |= (1ULL << (VTD_PAGE_SHIFT + mask - 1)) - 1;
1498 		desc.qw1 = QI_DEV_IOTLB_ADDR(addr) | QI_DEV_IOTLB_SIZE;
1499 	} else
1500 		desc.qw1 = QI_DEV_IOTLB_ADDR(addr);
1501 
1502 	if (qdep >= QI_DEV_IOTLB_MAX_INVS)
1503 		qdep = 0;
1504 
1505 	desc.qw0 = QI_DEV_IOTLB_SID(sid) | QI_DEV_IOTLB_QDEP(qdep) |
1506 		   QI_DIOTLB_TYPE | QI_DEV_IOTLB_PFSID(pfsid);
1507 	desc.qw2 = 0;
1508 	desc.qw3 = 0;
1509 
1510 	qi_submit_sync(iommu, &desc, 1, 0);
1511 }
1512 
1513 /* PASID-based IOTLB invalidation */
1514 void qi_flush_piotlb(struct intel_iommu *iommu, u16 did, u32 pasid, u64 addr,
1515 		     unsigned long npages, bool ih)
1516 {
1517 	struct qi_desc desc = {.qw2 = 0, .qw3 = 0};
1518 
1519 	/*
1520 	 * npages == -1 means a PASID-selective invalidation, otherwise,
1521 	 * a positive value for Page-selective-within-PASID invalidation.
1522 	 * 0 is not a valid input.
1523 	 */
1524 	if (WARN_ON(!npages)) {
1525 		pr_err("Invalid input npages = %ld\n", npages);
1526 		return;
1527 	}
1528 
1529 	if (npages == -1) {
1530 		desc.qw0 = QI_EIOTLB_PASID(pasid) |
1531 				QI_EIOTLB_DID(did) |
1532 				QI_EIOTLB_GRAN(QI_GRAN_NONG_PASID) |
1533 				QI_EIOTLB_TYPE;
1534 		desc.qw1 = 0;
1535 	} else {
1536 		int mask = ilog2(__roundup_pow_of_two(npages));
1537 		unsigned long align = (1ULL << (VTD_PAGE_SHIFT + mask));
1538 
1539 		if (WARN_ON_ONCE(!IS_ALIGNED(addr, align)))
1540 			addr = ALIGN_DOWN(addr, align);
1541 
1542 		desc.qw0 = QI_EIOTLB_PASID(pasid) |
1543 				QI_EIOTLB_DID(did) |
1544 				QI_EIOTLB_GRAN(QI_GRAN_PSI_PASID) |
1545 				QI_EIOTLB_TYPE;
1546 		desc.qw1 = QI_EIOTLB_ADDR(addr) |
1547 				QI_EIOTLB_IH(ih) |
1548 				QI_EIOTLB_AM(mask);
1549 	}
1550 
1551 	qi_submit_sync(iommu, &desc, 1, 0);
1552 }
1553 
1554 /* PASID-based device IOTLB Invalidate */
1555 void qi_flush_dev_iotlb_pasid(struct intel_iommu *iommu, u16 sid, u16 pfsid,
1556 			      u32 pasid,  u16 qdep, u64 addr, unsigned int size_order)
1557 {
1558 	unsigned long mask = 1UL << (VTD_PAGE_SHIFT + size_order - 1);
1559 	struct qi_desc desc = {.qw1 = 0, .qw2 = 0, .qw3 = 0};
1560 
1561 	desc.qw0 = QI_DEV_EIOTLB_PASID(pasid) | QI_DEV_EIOTLB_SID(sid) |
1562 		QI_DEV_EIOTLB_QDEP(qdep) | QI_DEIOTLB_TYPE |
1563 		QI_DEV_IOTLB_PFSID(pfsid);
1564 
1565 	/*
1566 	 * If S bit is 0, we only flush a single page. If S bit is set,
1567 	 * The least significant zero bit indicates the invalidation address
1568 	 * range. VT-d spec 6.5.2.6.
1569 	 * e.g. address bit 12[0] indicates 8KB, 13[0] indicates 16KB.
1570 	 * size order = 0 is PAGE_SIZE 4KB
1571 	 * Max Invs Pending (MIP) is set to 0 for now until we have DIT in
1572 	 * ECAP.
1573 	 */
1574 	if (!IS_ALIGNED(addr, VTD_PAGE_SIZE << size_order))
1575 		pr_warn_ratelimited("Invalidate non-aligned address %llx, order %d\n",
1576 				    addr, size_order);
1577 
1578 	/* Take page address */
1579 	desc.qw1 = QI_DEV_EIOTLB_ADDR(addr);
1580 
1581 	if (size_order) {
1582 		/*
1583 		 * Existing 0s in address below size_order may be the least
1584 		 * significant bit, we must set them to 1s to avoid having
1585 		 * smaller size than desired.
1586 		 */
1587 		desc.qw1 |= GENMASK_ULL(size_order + VTD_PAGE_SHIFT - 1,
1588 					VTD_PAGE_SHIFT);
1589 		/* Clear size_order bit to indicate size */
1590 		desc.qw1 &= ~mask;
1591 		/* Set the S bit to indicate flushing more than 1 page */
1592 		desc.qw1 |= QI_DEV_EIOTLB_SIZE;
1593 	}
1594 
1595 	qi_submit_sync(iommu, &desc, 1, 0);
1596 }
1597 
1598 void qi_flush_pasid_cache(struct intel_iommu *iommu, u16 did,
1599 			  u64 granu, u32 pasid)
1600 {
1601 	struct qi_desc desc = {.qw1 = 0, .qw2 = 0, .qw3 = 0};
1602 
1603 	desc.qw0 = QI_PC_PASID(pasid) | QI_PC_DID(did) |
1604 			QI_PC_GRAN(granu) | QI_PC_TYPE;
1605 	qi_submit_sync(iommu, &desc, 1, 0);
1606 }
1607 
1608 /*
1609  * Disable Queued Invalidation interface.
1610  */
1611 void dmar_disable_qi(struct intel_iommu *iommu)
1612 {
1613 	unsigned long flags;
1614 	u32 sts;
1615 	cycles_t start_time = get_cycles();
1616 
1617 	if (!ecap_qis(iommu->ecap))
1618 		return;
1619 
1620 	raw_spin_lock_irqsave(&iommu->register_lock, flags);
1621 
1622 	sts =  readl(iommu->reg + DMAR_GSTS_REG);
1623 	if (!(sts & DMA_GSTS_QIES))
1624 		goto end;
1625 
1626 	/*
1627 	 * Give a chance to HW to complete the pending invalidation requests.
1628 	 */
1629 	while ((readl(iommu->reg + DMAR_IQT_REG) !=
1630 		readl(iommu->reg + DMAR_IQH_REG)) &&
1631 		(DMAR_OPERATION_TIMEOUT > (get_cycles() - start_time)))
1632 		cpu_relax();
1633 
1634 	iommu->gcmd &= ~DMA_GCMD_QIE;
1635 	writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
1636 
1637 	IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl,
1638 		      !(sts & DMA_GSTS_QIES), sts);
1639 end:
1640 	raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
1641 }
1642 
1643 /*
1644  * Enable queued invalidation.
1645  */
1646 static void __dmar_enable_qi(struct intel_iommu *iommu)
1647 {
1648 	u32 sts;
1649 	unsigned long flags;
1650 	struct q_inval *qi = iommu->qi;
1651 	u64 val = virt_to_phys(qi->desc);
1652 
1653 	qi->free_head = qi->free_tail = 0;
1654 	qi->free_cnt = QI_LENGTH;
1655 
1656 	/*
1657 	 * Set DW=1 and QS=1 in IQA_REG when Scalable Mode capability
1658 	 * is present.
1659 	 */
1660 	if (ecap_smts(iommu->ecap))
1661 		val |= (1 << 11) | 1;
1662 
1663 	raw_spin_lock_irqsave(&iommu->register_lock, flags);
1664 
1665 	/* write zero to the tail reg */
1666 	writel(0, iommu->reg + DMAR_IQT_REG);
1667 
1668 	dmar_writeq(iommu->reg + DMAR_IQA_REG, val);
1669 
1670 	iommu->gcmd |= DMA_GCMD_QIE;
1671 	writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
1672 
1673 	/* Make sure hardware complete it */
1674 	IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_QIES), sts);
1675 
1676 	raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
1677 }
1678 
1679 /*
1680  * Enable Queued Invalidation interface. This is a must to support
1681  * interrupt-remapping. Also used by DMA-remapping, which replaces
1682  * register based IOTLB invalidation.
1683  */
1684 int dmar_enable_qi(struct intel_iommu *iommu)
1685 {
1686 	struct q_inval *qi;
1687 	struct page *desc_page;
1688 
1689 	if (!ecap_qis(iommu->ecap))
1690 		return -ENOENT;
1691 
1692 	/*
1693 	 * queued invalidation is already setup and enabled.
1694 	 */
1695 	if (iommu->qi)
1696 		return 0;
1697 
1698 	iommu->qi = kmalloc(sizeof(*qi), GFP_ATOMIC);
1699 	if (!iommu->qi)
1700 		return -ENOMEM;
1701 
1702 	qi = iommu->qi;
1703 
1704 	/*
1705 	 * Need two pages to accommodate 256 descriptors of 256 bits each
1706 	 * if the remapping hardware supports scalable mode translation.
1707 	 */
1708 	desc_page = alloc_pages_node(iommu->node, GFP_ATOMIC | __GFP_ZERO,
1709 				     !!ecap_smts(iommu->ecap));
1710 	if (!desc_page) {
1711 		kfree(qi);
1712 		iommu->qi = NULL;
1713 		return -ENOMEM;
1714 	}
1715 
1716 	qi->desc = page_address(desc_page);
1717 
1718 	qi->desc_status = kcalloc(QI_LENGTH, sizeof(int), GFP_ATOMIC);
1719 	if (!qi->desc_status) {
1720 		free_page((unsigned long) qi->desc);
1721 		kfree(qi);
1722 		iommu->qi = NULL;
1723 		return -ENOMEM;
1724 	}
1725 
1726 	raw_spin_lock_init(&qi->q_lock);
1727 
1728 	__dmar_enable_qi(iommu);
1729 
1730 	return 0;
1731 }
1732 
1733 /* iommu interrupt handling. Most stuff are MSI-like. */
1734 
1735 enum faulttype {
1736 	DMA_REMAP,
1737 	INTR_REMAP,
1738 	UNKNOWN,
1739 };
1740 
1741 static const char *dma_remap_fault_reasons[] =
1742 {
1743 	"Software",
1744 	"Present bit in root entry is clear",
1745 	"Present bit in context entry is clear",
1746 	"Invalid context entry",
1747 	"Access beyond MGAW",
1748 	"PTE Write access is not set",
1749 	"PTE Read access is not set",
1750 	"Next page table ptr is invalid",
1751 	"Root table address invalid",
1752 	"Context table ptr is invalid",
1753 	"non-zero reserved fields in RTP",
1754 	"non-zero reserved fields in CTP",
1755 	"non-zero reserved fields in PTE",
1756 	"PCE for translation request specifies blocking",
1757 };
1758 
1759 static const char * const dma_remap_sm_fault_reasons[] = {
1760 	"SM: Invalid Root Table Address",
1761 	"SM: TTM 0 for request with PASID",
1762 	"SM: TTM 0 for page group request",
1763 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x33-0x37 */
1764 	"SM: Error attempting to access Root Entry",
1765 	"SM: Present bit in Root Entry is clear",
1766 	"SM: Non-zero reserved field set in Root Entry",
1767 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x3B-0x3F */
1768 	"SM: Error attempting to access Context Entry",
1769 	"SM: Present bit in Context Entry is clear",
1770 	"SM: Non-zero reserved field set in the Context Entry",
1771 	"SM: Invalid Context Entry",
1772 	"SM: DTE field in Context Entry is clear",
1773 	"SM: PASID Enable field in Context Entry is clear",
1774 	"SM: PASID is larger than the max in Context Entry",
1775 	"SM: PRE field in Context-Entry is clear",
1776 	"SM: RID_PASID field error in Context-Entry",
1777 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x49-0x4F */
1778 	"SM: Error attempting to access the PASID Directory Entry",
1779 	"SM: Present bit in Directory Entry is clear",
1780 	"SM: Non-zero reserved field set in PASID Directory Entry",
1781 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x53-0x57 */
1782 	"SM: Error attempting to access PASID Table Entry",
1783 	"SM: Present bit in PASID Table Entry is clear",
1784 	"SM: Non-zero reserved field set in PASID Table Entry",
1785 	"SM: Invalid Scalable-Mode PASID Table Entry",
1786 	"SM: ERE field is clear in PASID Table Entry",
1787 	"SM: SRE field is clear in PASID Table Entry",
1788 	"Unknown", "Unknown",/* 0x5E-0x5F */
1789 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x60-0x67 */
1790 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x68-0x6F */
1791 	"SM: Error attempting to access first-level paging entry",
1792 	"SM: Present bit in first-level paging entry is clear",
1793 	"SM: Non-zero reserved field set in first-level paging entry",
1794 	"SM: Error attempting to access FL-PML4 entry",
1795 	"SM: First-level entry address beyond MGAW in Nested translation",
1796 	"SM: Read permission error in FL-PML4 entry in Nested translation",
1797 	"SM: Read permission error in first-level paging entry in Nested translation",
1798 	"SM: Write permission error in first-level paging entry in Nested translation",
1799 	"SM: Error attempting to access second-level paging entry",
1800 	"SM: Read/Write permission error in second-level paging entry",
1801 	"SM: Non-zero reserved field set in second-level paging entry",
1802 	"SM: Invalid second-level page table pointer",
1803 	"SM: A/D bit update needed in second-level entry when set up in no snoop",
1804 	"Unknown", "Unknown", "Unknown", /* 0x7D-0x7F */
1805 	"SM: Address in first-level translation is not canonical",
1806 	"SM: U/S set 0 for first-level translation with user privilege",
1807 	"SM: No execute permission for request with PASID and ER=1",
1808 	"SM: Address beyond the DMA hardware max",
1809 	"SM: Second-level entry address beyond the max",
1810 	"SM: No write permission for Write/AtomicOp request",
1811 	"SM: No read permission for Read/AtomicOp request",
1812 	"SM: Invalid address-interrupt address",
1813 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x88-0x8F */
1814 	"SM: A/D bit update needed in first-level entry when set up in no snoop",
1815 };
1816 
1817 static const char *irq_remap_fault_reasons[] =
1818 {
1819 	"Detected reserved fields in the decoded interrupt-remapped request",
1820 	"Interrupt index exceeded the interrupt-remapping table size",
1821 	"Present field in the IRTE entry is clear",
1822 	"Error accessing interrupt-remapping table pointed by IRTA_REG",
1823 	"Detected reserved fields in the IRTE entry",
1824 	"Blocked a compatibility format interrupt request",
1825 	"Blocked an interrupt request due to source-id verification failure",
1826 };
1827 
1828 static const char *dmar_get_fault_reason(u8 fault_reason, int *fault_type)
1829 {
1830 	if (fault_reason >= 0x20 && (fault_reason - 0x20 <
1831 					ARRAY_SIZE(irq_remap_fault_reasons))) {
1832 		*fault_type = INTR_REMAP;
1833 		return irq_remap_fault_reasons[fault_reason - 0x20];
1834 	} else if (fault_reason >= 0x30 && (fault_reason - 0x30 <
1835 			ARRAY_SIZE(dma_remap_sm_fault_reasons))) {
1836 		*fault_type = DMA_REMAP;
1837 		return dma_remap_sm_fault_reasons[fault_reason - 0x30];
1838 	} else if (fault_reason < ARRAY_SIZE(dma_remap_fault_reasons)) {
1839 		*fault_type = DMA_REMAP;
1840 		return dma_remap_fault_reasons[fault_reason];
1841 	} else {
1842 		*fault_type = UNKNOWN;
1843 		return "Unknown";
1844 	}
1845 }
1846 
1847 
1848 static inline int dmar_msi_reg(struct intel_iommu *iommu, int irq)
1849 {
1850 	if (iommu->irq == irq)
1851 		return DMAR_FECTL_REG;
1852 	else if (iommu->pr_irq == irq)
1853 		return DMAR_PECTL_REG;
1854 	else
1855 		BUG();
1856 }
1857 
1858 void dmar_msi_unmask(struct irq_data *data)
1859 {
1860 	struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
1861 	int reg = dmar_msi_reg(iommu, data->irq);
1862 	unsigned long flag;
1863 
1864 	/* unmask it */
1865 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1866 	writel(0, iommu->reg + reg);
1867 	/* Read a reg to force flush the post write */
1868 	readl(iommu->reg + reg);
1869 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1870 }
1871 
1872 void dmar_msi_mask(struct irq_data *data)
1873 {
1874 	struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
1875 	int reg = dmar_msi_reg(iommu, data->irq);
1876 	unsigned long flag;
1877 
1878 	/* mask it */
1879 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1880 	writel(DMA_FECTL_IM, iommu->reg + reg);
1881 	/* Read a reg to force flush the post write */
1882 	readl(iommu->reg + reg);
1883 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1884 }
1885 
1886 void dmar_msi_write(int irq, struct msi_msg *msg)
1887 {
1888 	struct intel_iommu *iommu = irq_get_handler_data(irq);
1889 	int reg = dmar_msi_reg(iommu, irq);
1890 	unsigned long flag;
1891 
1892 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1893 	writel(msg->data, iommu->reg + reg + 4);
1894 	writel(msg->address_lo, iommu->reg + reg + 8);
1895 	writel(msg->address_hi, iommu->reg + reg + 12);
1896 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1897 }
1898 
1899 void dmar_msi_read(int irq, struct msi_msg *msg)
1900 {
1901 	struct intel_iommu *iommu = irq_get_handler_data(irq);
1902 	int reg = dmar_msi_reg(iommu, irq);
1903 	unsigned long flag;
1904 
1905 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1906 	msg->data = readl(iommu->reg + reg + 4);
1907 	msg->address_lo = readl(iommu->reg + reg + 8);
1908 	msg->address_hi = readl(iommu->reg + reg + 12);
1909 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1910 }
1911 
1912 static int dmar_fault_do_one(struct intel_iommu *iommu, int type,
1913 		u8 fault_reason, u32 pasid, u16 source_id,
1914 		unsigned long long addr)
1915 {
1916 	const char *reason;
1917 	int fault_type;
1918 
1919 	reason = dmar_get_fault_reason(fault_reason, &fault_type);
1920 
1921 	if (fault_type == INTR_REMAP) {
1922 		pr_err("[INTR-REMAP] Request device [%02x:%02x.%d] fault index 0x%llx [fault reason 0x%02x] %s\n",
1923 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1924 		       PCI_FUNC(source_id & 0xFF), addr >> 48,
1925 		       fault_reason, reason);
1926 
1927 		return 0;
1928 	}
1929 
1930 	if (pasid == INVALID_IOASID)
1931 		pr_err("[%s NO_PASID] Request device [%02x:%02x.%d] fault addr 0x%llx [fault reason 0x%02x] %s\n",
1932 		       type ? "DMA Read" : "DMA Write",
1933 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1934 		       PCI_FUNC(source_id & 0xFF), addr,
1935 		       fault_reason, reason);
1936 	else
1937 		pr_err("[%s PASID 0x%x] Request device [%02x:%02x.%d] fault addr 0x%llx [fault reason 0x%02x] %s\n",
1938 		       type ? "DMA Read" : "DMA Write", pasid,
1939 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1940 		       PCI_FUNC(source_id & 0xFF), addr,
1941 		       fault_reason, reason);
1942 
1943 	dmar_fault_dump_ptes(iommu, source_id, addr, pasid);
1944 
1945 	return 0;
1946 }
1947 
1948 #define PRIMARY_FAULT_REG_LEN (16)
1949 irqreturn_t dmar_fault(int irq, void *dev_id)
1950 {
1951 	struct intel_iommu *iommu = dev_id;
1952 	int reg, fault_index;
1953 	u32 fault_status;
1954 	unsigned long flag;
1955 	static DEFINE_RATELIMIT_STATE(rs,
1956 				      DEFAULT_RATELIMIT_INTERVAL,
1957 				      DEFAULT_RATELIMIT_BURST);
1958 
1959 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1960 	fault_status = readl(iommu->reg + DMAR_FSTS_REG);
1961 	if (fault_status && __ratelimit(&rs))
1962 		pr_err("DRHD: handling fault status reg %x\n", fault_status);
1963 
1964 	/* TBD: ignore advanced fault log currently */
1965 	if (!(fault_status & DMA_FSTS_PPF))
1966 		goto unlock_exit;
1967 
1968 	fault_index = dma_fsts_fault_record_index(fault_status);
1969 	reg = cap_fault_reg_offset(iommu->cap);
1970 	while (1) {
1971 		/* Disable printing, simply clear the fault when ratelimited */
1972 		bool ratelimited = !__ratelimit(&rs);
1973 		u8 fault_reason;
1974 		u16 source_id;
1975 		u64 guest_addr;
1976 		u32 pasid;
1977 		int type;
1978 		u32 data;
1979 		bool pasid_present;
1980 
1981 		/* highest 32 bits */
1982 		data = readl(iommu->reg + reg +
1983 				fault_index * PRIMARY_FAULT_REG_LEN + 12);
1984 		if (!(data & DMA_FRCD_F))
1985 			break;
1986 
1987 		if (!ratelimited) {
1988 			fault_reason = dma_frcd_fault_reason(data);
1989 			type = dma_frcd_type(data);
1990 
1991 			pasid = dma_frcd_pasid_value(data);
1992 			data = readl(iommu->reg + reg +
1993 				     fault_index * PRIMARY_FAULT_REG_LEN + 8);
1994 			source_id = dma_frcd_source_id(data);
1995 
1996 			pasid_present = dma_frcd_pasid_present(data);
1997 			guest_addr = dmar_readq(iommu->reg + reg +
1998 					fault_index * PRIMARY_FAULT_REG_LEN);
1999 			guest_addr = dma_frcd_page_addr(guest_addr);
2000 		}
2001 
2002 		/* clear the fault */
2003 		writel(DMA_FRCD_F, iommu->reg + reg +
2004 			fault_index * PRIMARY_FAULT_REG_LEN + 12);
2005 
2006 		raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
2007 
2008 		if (!ratelimited)
2009 			/* Using pasid -1 if pasid is not present */
2010 			dmar_fault_do_one(iommu, type, fault_reason,
2011 					  pasid_present ? pasid : INVALID_IOASID,
2012 					  source_id, guest_addr);
2013 
2014 		fault_index++;
2015 		if (fault_index >= cap_num_fault_regs(iommu->cap))
2016 			fault_index = 0;
2017 		raw_spin_lock_irqsave(&iommu->register_lock, flag);
2018 	}
2019 
2020 	writel(DMA_FSTS_PFO | DMA_FSTS_PPF | DMA_FSTS_PRO,
2021 	       iommu->reg + DMAR_FSTS_REG);
2022 
2023 unlock_exit:
2024 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
2025 	return IRQ_HANDLED;
2026 }
2027 
2028 int dmar_set_interrupt(struct intel_iommu *iommu)
2029 {
2030 	int irq, ret;
2031 
2032 	/*
2033 	 * Check if the fault interrupt is already initialized.
2034 	 */
2035 	if (iommu->irq)
2036 		return 0;
2037 
2038 	irq = dmar_alloc_hwirq(iommu->seq_id, iommu->node, iommu);
2039 	if (irq > 0) {
2040 		iommu->irq = irq;
2041 	} else {
2042 		pr_err("No free IRQ vectors\n");
2043 		return -EINVAL;
2044 	}
2045 
2046 	ret = request_irq(irq, dmar_fault, IRQF_NO_THREAD, iommu->name, iommu);
2047 	if (ret)
2048 		pr_err("Can't request irq\n");
2049 	return ret;
2050 }
2051 
2052 int __init enable_drhd_fault_handling(void)
2053 {
2054 	struct dmar_drhd_unit *drhd;
2055 	struct intel_iommu *iommu;
2056 
2057 	/*
2058 	 * Enable fault control interrupt.
2059 	 */
2060 	for_each_iommu(iommu, drhd) {
2061 		u32 fault_status;
2062 		int ret = dmar_set_interrupt(iommu);
2063 
2064 		if (ret) {
2065 			pr_err("DRHD %Lx: failed to enable fault, interrupt, ret %d\n",
2066 			       (unsigned long long)drhd->reg_base_addr, ret);
2067 			return -1;
2068 		}
2069 
2070 		/*
2071 		 * Clear any previous faults.
2072 		 */
2073 		dmar_fault(iommu->irq, iommu);
2074 		fault_status = readl(iommu->reg + DMAR_FSTS_REG);
2075 		writel(fault_status, iommu->reg + DMAR_FSTS_REG);
2076 	}
2077 
2078 	return 0;
2079 }
2080 
2081 /*
2082  * Re-enable Queued Invalidation interface.
2083  */
2084 int dmar_reenable_qi(struct intel_iommu *iommu)
2085 {
2086 	if (!ecap_qis(iommu->ecap))
2087 		return -ENOENT;
2088 
2089 	if (!iommu->qi)
2090 		return -ENOENT;
2091 
2092 	/*
2093 	 * First disable queued invalidation.
2094 	 */
2095 	dmar_disable_qi(iommu);
2096 	/*
2097 	 * Then enable queued invalidation again. Since there is no pending
2098 	 * invalidation requests now, it's safe to re-enable queued
2099 	 * invalidation.
2100 	 */
2101 	__dmar_enable_qi(iommu);
2102 
2103 	return 0;
2104 }
2105 
2106 /*
2107  * Check interrupt remapping support in DMAR table description.
2108  */
2109 int __init dmar_ir_support(void)
2110 {
2111 	struct acpi_table_dmar *dmar;
2112 	dmar = (struct acpi_table_dmar *)dmar_tbl;
2113 	if (!dmar)
2114 		return 0;
2115 	return dmar->flags & 0x1;
2116 }
2117 
2118 /* Check whether DMAR units are in use */
2119 static inline bool dmar_in_use(void)
2120 {
2121 	return irq_remapping_enabled || intel_iommu_enabled;
2122 }
2123 
2124 static int __init dmar_free_unused_resources(void)
2125 {
2126 	struct dmar_drhd_unit *dmaru, *dmaru_n;
2127 
2128 	if (dmar_in_use())
2129 		return 0;
2130 
2131 	if (dmar_dev_scope_status != 1 && !list_empty(&dmar_drhd_units))
2132 		bus_unregister_notifier(&pci_bus_type, &dmar_pci_bus_nb);
2133 
2134 	down_write(&dmar_global_lock);
2135 	list_for_each_entry_safe(dmaru, dmaru_n, &dmar_drhd_units, list) {
2136 		list_del(&dmaru->list);
2137 		dmar_free_drhd(dmaru);
2138 	}
2139 	up_write(&dmar_global_lock);
2140 
2141 	return 0;
2142 }
2143 
2144 late_initcall(dmar_free_unused_resources);
2145 
2146 /*
2147  * DMAR Hotplug Support
2148  * For more details, please refer to Intel(R) Virtualization Technology
2149  * for Directed-IO Architecture Specifiction, Rev 2.2, Section 8.8
2150  * "Remapping Hardware Unit Hot Plug".
2151  */
2152 static guid_t dmar_hp_guid =
2153 	GUID_INIT(0xD8C1A3A6, 0xBE9B, 0x4C9B,
2154 		  0x91, 0xBF, 0xC3, 0xCB, 0x81, 0xFC, 0x5D, 0xAF);
2155 
2156 /*
2157  * Currently there's only one revision and BIOS will not check the revision id,
2158  * so use 0 for safety.
2159  */
2160 #define	DMAR_DSM_REV_ID			0
2161 #define	DMAR_DSM_FUNC_DRHD		1
2162 #define	DMAR_DSM_FUNC_ATSR		2
2163 #define	DMAR_DSM_FUNC_RHSA		3
2164 #define	DMAR_DSM_FUNC_SATC		4
2165 
2166 static inline bool dmar_detect_dsm(acpi_handle handle, int func)
2167 {
2168 	return acpi_check_dsm(handle, &dmar_hp_guid, DMAR_DSM_REV_ID, 1 << func);
2169 }
2170 
2171 static int dmar_walk_dsm_resource(acpi_handle handle, int func,
2172 				  dmar_res_handler_t handler, void *arg)
2173 {
2174 	int ret = -ENODEV;
2175 	union acpi_object *obj;
2176 	struct acpi_dmar_header *start;
2177 	struct dmar_res_callback callback;
2178 	static int res_type[] = {
2179 		[DMAR_DSM_FUNC_DRHD] = ACPI_DMAR_TYPE_HARDWARE_UNIT,
2180 		[DMAR_DSM_FUNC_ATSR] = ACPI_DMAR_TYPE_ROOT_ATS,
2181 		[DMAR_DSM_FUNC_RHSA] = ACPI_DMAR_TYPE_HARDWARE_AFFINITY,
2182 		[DMAR_DSM_FUNC_SATC] = ACPI_DMAR_TYPE_SATC,
2183 	};
2184 
2185 	if (!dmar_detect_dsm(handle, func))
2186 		return 0;
2187 
2188 	obj = acpi_evaluate_dsm_typed(handle, &dmar_hp_guid, DMAR_DSM_REV_ID,
2189 				      func, NULL, ACPI_TYPE_BUFFER);
2190 	if (!obj)
2191 		return -ENODEV;
2192 
2193 	memset(&callback, 0, sizeof(callback));
2194 	callback.cb[res_type[func]] = handler;
2195 	callback.arg[res_type[func]] = arg;
2196 	start = (struct acpi_dmar_header *)obj->buffer.pointer;
2197 	ret = dmar_walk_remapping_entries(start, obj->buffer.length, &callback);
2198 
2199 	ACPI_FREE(obj);
2200 
2201 	return ret;
2202 }
2203 
2204 static int dmar_hp_add_drhd(struct acpi_dmar_header *header, void *arg)
2205 {
2206 	int ret;
2207 	struct dmar_drhd_unit *dmaru;
2208 
2209 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2210 	if (!dmaru)
2211 		return -ENODEV;
2212 
2213 	ret = dmar_ir_hotplug(dmaru, true);
2214 	if (ret == 0)
2215 		ret = dmar_iommu_hotplug(dmaru, true);
2216 
2217 	return ret;
2218 }
2219 
2220 static int dmar_hp_remove_drhd(struct acpi_dmar_header *header, void *arg)
2221 {
2222 	int i, ret;
2223 	struct device *dev;
2224 	struct dmar_drhd_unit *dmaru;
2225 
2226 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2227 	if (!dmaru)
2228 		return 0;
2229 
2230 	/*
2231 	 * All PCI devices managed by this unit should have been destroyed.
2232 	 */
2233 	if (!dmaru->include_all && dmaru->devices && dmaru->devices_cnt) {
2234 		for_each_active_dev_scope(dmaru->devices,
2235 					  dmaru->devices_cnt, i, dev)
2236 			return -EBUSY;
2237 	}
2238 
2239 	ret = dmar_ir_hotplug(dmaru, false);
2240 	if (ret == 0)
2241 		ret = dmar_iommu_hotplug(dmaru, false);
2242 
2243 	return ret;
2244 }
2245 
2246 static int dmar_hp_release_drhd(struct acpi_dmar_header *header, void *arg)
2247 {
2248 	struct dmar_drhd_unit *dmaru;
2249 
2250 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2251 	if (dmaru) {
2252 		list_del_rcu(&dmaru->list);
2253 		synchronize_rcu();
2254 		dmar_free_drhd(dmaru);
2255 	}
2256 
2257 	return 0;
2258 }
2259 
2260 static int dmar_hotplug_insert(acpi_handle handle)
2261 {
2262 	int ret;
2263 	int drhd_count = 0;
2264 
2265 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2266 				     &dmar_validate_one_drhd, (void *)1);
2267 	if (ret)
2268 		goto out;
2269 
2270 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2271 				     &dmar_parse_one_drhd, (void *)&drhd_count);
2272 	if (ret == 0 && drhd_count == 0) {
2273 		pr_warn(FW_BUG "No DRHD structures in buffer returned by _DSM method\n");
2274 		goto out;
2275 	} else if (ret) {
2276 		goto release_drhd;
2277 	}
2278 
2279 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_RHSA,
2280 				     &dmar_parse_one_rhsa, NULL);
2281 	if (ret)
2282 		goto release_drhd;
2283 
2284 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2285 				     &dmar_parse_one_atsr, NULL);
2286 	if (ret)
2287 		goto release_atsr;
2288 
2289 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2290 				     &dmar_hp_add_drhd, NULL);
2291 	if (!ret)
2292 		return 0;
2293 
2294 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2295 			       &dmar_hp_remove_drhd, NULL);
2296 release_atsr:
2297 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2298 			       &dmar_release_one_atsr, NULL);
2299 release_drhd:
2300 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2301 			       &dmar_hp_release_drhd, NULL);
2302 out:
2303 	return ret;
2304 }
2305 
2306 static int dmar_hotplug_remove(acpi_handle handle)
2307 {
2308 	int ret;
2309 
2310 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2311 				     &dmar_check_one_atsr, NULL);
2312 	if (ret)
2313 		return ret;
2314 
2315 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2316 				     &dmar_hp_remove_drhd, NULL);
2317 	if (ret == 0) {
2318 		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2319 					       &dmar_release_one_atsr, NULL));
2320 		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2321 					       &dmar_hp_release_drhd, NULL));
2322 	} else {
2323 		dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2324 				       &dmar_hp_add_drhd, NULL);
2325 	}
2326 
2327 	return ret;
2328 }
2329 
2330 static acpi_status dmar_get_dsm_handle(acpi_handle handle, u32 lvl,
2331 				       void *context, void **retval)
2332 {
2333 	acpi_handle *phdl = retval;
2334 
2335 	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
2336 		*phdl = handle;
2337 		return AE_CTRL_TERMINATE;
2338 	}
2339 
2340 	return AE_OK;
2341 }
2342 
2343 static int dmar_device_hotplug(acpi_handle handle, bool insert)
2344 {
2345 	int ret;
2346 	acpi_handle tmp = NULL;
2347 	acpi_status status;
2348 
2349 	if (!dmar_in_use())
2350 		return 0;
2351 
2352 	/*
2353 	 * It's unlikely that any I/O board is hot added before the IOMMU
2354 	 * subsystem is initialized.
2355 	 */
2356 	if (IS_ENABLED(CONFIG_INTEL_IOMMU) && !intel_iommu_enabled)
2357 		return -EOPNOTSUPP;
2358 
2359 	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
2360 		tmp = handle;
2361 	} else {
2362 		status = acpi_walk_namespace(ACPI_TYPE_DEVICE, handle,
2363 					     ACPI_UINT32_MAX,
2364 					     dmar_get_dsm_handle,
2365 					     NULL, NULL, &tmp);
2366 		if (ACPI_FAILURE(status)) {
2367 			pr_warn("Failed to locate _DSM method.\n");
2368 			return -ENXIO;
2369 		}
2370 	}
2371 	if (tmp == NULL)
2372 		return 0;
2373 
2374 	down_write(&dmar_global_lock);
2375 	if (insert)
2376 		ret = dmar_hotplug_insert(tmp);
2377 	else
2378 		ret = dmar_hotplug_remove(tmp);
2379 	up_write(&dmar_global_lock);
2380 
2381 	return ret;
2382 }
2383 
2384 int dmar_device_add(acpi_handle handle)
2385 {
2386 	return dmar_device_hotplug(handle, true);
2387 }
2388 
2389 int dmar_device_remove(acpi_handle handle)
2390 {
2391 	return dmar_device_hotplug(handle, false);
2392 }
2393 
2394 /*
2395  * dmar_platform_optin - Is %DMA_CTRL_PLATFORM_OPT_IN_FLAG set in DMAR table
2396  *
2397  * Returns true if the platform has %DMA_CTRL_PLATFORM_OPT_IN_FLAG set in
2398  * the ACPI DMAR table. This means that the platform boot firmware has made
2399  * sure no device can issue DMA outside of RMRR regions.
2400  */
2401 bool dmar_platform_optin(void)
2402 {
2403 	struct acpi_table_dmar *dmar;
2404 	acpi_status status;
2405 	bool ret;
2406 
2407 	status = acpi_get_table(ACPI_SIG_DMAR, 0,
2408 				(struct acpi_table_header **)&dmar);
2409 	if (ACPI_FAILURE(status))
2410 		return false;
2411 
2412 	ret = !!(dmar->flags & DMAR_PLATFORM_OPT_IN);
2413 	acpi_put_table((struct acpi_table_header *)dmar);
2414 
2415 	return ret;
2416 }
2417 EXPORT_SYMBOL_GPL(dmar_platform_optin);
2418